January 25th, 2014

Alice in Puzzle-Land

The Wonderland-themed Mystery Hunt was an idea that was bound to happen eventually. It came in second to the Mario theme on the final ballot in 2011, and in 2012 the proposal survived long enough to be the theme of a meta; I'm glad it was finally put to use as a full Hunt theme in such a successful Hunt. Many compliments to Alice Shrugged (great name, by the way) for pulling it off in style; this year's Mystery Hunt was amazingly clean and elegant with high-quality puzzles all around. I don't think I really felt blown away by awesomeness about any of the individual puzzles, as I sometimes have even in Hunts that were overall quite problematic; but the baseline level of puzzle quality was so high that I scarcely would consider this a criticism.

Alice Shrugged pulled off some feats that are really hard to accomplish; my hat is off to them. First and foremost, every single puzzle, including all the metas, was solved by at least seven teams, which indicates both an amazingly high level of quality control and puzzle difficulty calibration and also fairly clever (or lucky) puzzle organization balancing (for instance, in 2011 we had a puzzle that no team solved because it was inadequately test-solved, but we also had a puzzle that no team solved simply because it was long and late in the Hunt). This might be the single most impressive thing about this Hunt.

Almost as impressive, however, and perhaps more unique, is the following: they got event puzzles right, which I'm not sure any other Mystery Hunt has accomplished since event puzzles really became a thing. The one I went to (the singing event) was fun, got everyone involved, had a meaningful role in Hunt structure, and didn't run too long, and from the reactions of my teammates it seems that the same was true for all the other events as well. Having seen many failed events both from the Hunt-writing side and from the solving side, I feel safe saying this is no mean feat and future teams could learn a thing or two from how events worked this year.

This is something I care more about than a lot of people on my team, but I was also really pleased overall with how well and how often the answers matched the puzzle topics. There were multiple times when we solved a puzzle and extracted the answer and I was like 'Well, of course the answer is ____, that's what the whole puzzle has been about!"—not that I could have guessed the answer without solving the puzzle, but merely that the answer seemed related to the theme or mechanism of the puzzle in a natural way. I always feel that makes a puzzle seem like a more cohesive unit, rather than just a way of getting you one more input to the metapuzzle.

I only have one and a half real criticisms of this Hunt. The half criticism is that the kickoff segue from the fake "academic conference" theme to the actual Alice theme was kind of abrupt and awkward; I don't really think they needed a fake theme at all. I call that a half criticism because it's obviously really trivial.

The real criticism is that I feel the unlocking of new puzzles was too restrictive. I don't mean that in the sense that too few puzzles were unlocked at a time—members of Alice Shrugged have explained that they wanted to keep unlock speed fairly low to avoid overwhelming smaller teams, which is perfectly reasonable—but rather that which puzzles unlocked which other puzzles seemed too restrictive. As far as I understand the unlocking in this Hunt, once you had unlocked a round, further puzzles in that round could only be unlocked by solving puzzles in that round. This type of unlocking is highly susceptible to bottlenecks—Plant was stuck on the Knights round for quite a while, while solving puzzles in other rounds at a decent clip, and being stuck on the Knights puzzles that we had meant we had no way of getting more Knight puzzles to try and solve around them (until a timed release finally opened the rest of the round for us). In 2011, we deliberately avoided bottlenecks; the order of unlocking puzzles was deterministic and any solve contributed to your total points for whatever the next unlock was. That might have resulted in the structure of the Hunt feeling less exploratory; but it made it a lot harder to get stuck somewhere.

Plant came in eighth this year—like in 2012, we were the last team to complete endgame before Hunt operations officially closed down. That means we got to complete the Hunt without working any harder than we had to, which was more or less our goal. It's interesting to see how the conversation on the team has evolved from our post-2011 'LET'S NEVER WIN AGAIN' to 'well, if we won again…' to 'well, maybe we should win again sometime…' over the past couple of years… so I guess, don't count Plant out yet.

Thanks and salutations to Alice Shrugged for a great Hunt, and congratulations to the Random team!

In a later post, I'll probably comment on a few specific puzzles that I found noteworthy.